[Ardali]

Dear All,
I work with immersed boundary to develop new class of dynamic mesh.
The problem that I have is previously reported on the forum. The immersed boundary is not robust and with small changes in the *.stl file will crash.
The Fatal Error that I receive is mainly:
"Can't find nearest triSurface point for cell ... "

I tried all tutorials available in the latest versions, they working well, while small change in the geometry, like replacing the cylinder with an airfoil, or sphere with a vane leads to an Fatal Error.

I changed the solver, codes, radiusFactor and etc, the outcome was not interesting.

Is there anybody who has debugged the code and knows anything more than me!
Bests
Ardalan

[hxaxtma]

From my experience this is only caused by a too coarse background mesh. Just try to create a more densed mesh in your area of interest and try it again.

[Ardali]

Thanks for your reply,
As you know, "refineImmersedBoundaryMesh.H" is available and a utility can be written to make the resolution finer locally, or an other way, make it finer globally using "BlockMeshDict".

I went up to more than 2.5 million cells , but still it does not work!!
In which cases you succeeded to run the simulations? I mean how complex was your *.stl file.

For double checking, I generate *.stl file from surface mesh in ICEM and convert it to *.ftr whit surfaceConvert!
Bests
Ardalan

[hxaxtma]

If you send me your stl file and your blockMeshDict I will have a look

[Ardali]

Thanks,
Please download the case:
https://www.dropbox.com/s/fux08tj0gv...annel.zip?dl=0

I use constant/triSurface/body3D.stl
Ardalan

[Tobi]

Hi,

you are using immersedBoundary in which version? I think foamExtend. Last week in the PFAU in Vienna we had a talk about a new implementation of the immersedBoundary theory in OpenFOAM. The guys were from Graz (Austria). Maybe it is possible to get the source code. As far as I understood, they do not have any problems.

[hxaxtma]

Well I checked your case and it worked for me,
so maybe you are missing some step in the pre processing,

1) Convert your stl to ftr with surfaceFeatureEdges *.stl *.ftr -angle 30
2) Run blockMesh
3) Add
ibSphere
{
type immersedBoundary;
nFaces 0;
startFace 263100;
internalFlow no;
}

to boundary file, note startFace must be equal to your first BC when running blockMesh
4) As I already mentioned, your basemesh resolution near the IB Area is too low, so improve your blockMeshDict

5) Run potentialIbFoam
6) Run your Ib Solver

The picture below shows the refined IBCells with refineImmersedBoundary -ibCells, of course the resolution is still not adequat

[Ardali]

Hi Tobias,
Thank you.
Yes I use FOAM_Extended-4.0 and the latest release.
It sounds great if their code works fine.
Is there any possibility to have access to the code?
What is the problem with current code?
Bests
Ardalan

[hxaxtma]

Originally Posted by Tobi

Hi,

you are using immersedBoundary in which version? I think foamExtend. Last week in the PFAU in Vienna we had a talk about a new implementation of the immersedBoundary theory in OpenFOAM. The guys were from Graz (Austria). Maybe it is possible to get the source code. As far as I understood, they do not have any problems.

@Tobi,
who exactly pronounced the IB solver? I am highly interested in*g*

[Tobi]

Hi all,

I don't know if there is any problem in the IB from the extend project. However, Federico Municchi from TUGraz hold some presentation about Fictitious Domain and immersed Boundary methods in OpenFOAM.

[Ardali]

Originally Posted by hxaxtma

Well I checked your case and it worked for me,
so maybe you are missing some step in the pre processing,

1) Convert your stl to ftr with surfaceFeatureEdges *.stl *.ftr -angle 30
2) Run blockMesh
3) Add
ibSphere
{
type immersedBoundary;
nFaces 0;
startFace 263100;
internalFlow no;
}

to boundary file, note startFace must be equal to your first BC when running blockMesh
4) As I already mentioned, your basemesh resolution near the IB Area is too low, so improve your blockMeshDict

5) Run potentialIbFoam
6) Run your Ib Solver

The picture below shows the refined IBCells with refineImmersedBoundary -ibCells, of course the resolution is still not adequat

Thank you for your time,
Would you please send me the blockMeshDict that you made?
I could not run the case even with 4 times finer resolution.
Bests
Ardalan

[hxaxtma]

Here is your working case
1) run blockMesh
2) copy boundary from case folder to polyMesh folder
3) run refineImmersedBoundary -ibCells
4) run potentialIbFoam
5) Have fun!

[Ardali]

The problem was the compilation of Foam-extended-4.0. If you use the latest version you might have a problem with immersed boundary. Even some tutorials do not work properly. That is safer to stick to the previous version, Foam-Extended-3.2.
If you are going to use the latest version, there is an update released last week. consider it.
The problem is reported and discussed with administration!

Thank you all to replay and let us find the problem.
One more problem that still exists, is IbMask does not correspond with the geometry. Please follow the case shared in previous message.

Ardalan